Kii Three - my impressions and pro reviews

[Emlin]

Almost certainly. The DSP will stop overload. It's designed that way. How about you? You seem to get hot under the collar with little provocation, and take an inordinate time to cool down.

[Emlin]

That said, I did sense a limit of how loud the Kii Threes would play. Putzeys explained:

"The average output in digital recordings varies between –4dB for a lot of modern pop and rock and –20dB for some audiophile recordings, especially classical. What other high-end systems do is simply put in so much gain that you can play quiet recordings ridiculously loud but the risk of what happens when they forget to take the volume down when playing something else is borne completely by the user! I don't find that an acceptable solution. So I set the input sensitivity such that a maximally loud signal replayed at 100 is the loudest the speaker can do (allowing excursion limitation for sub frequencies, of course). Normally speaking, good recordings will be mastered such that at least the loudest passage will actually hit full scale, give or take, which means that 100 is still as loud as the speaker will go and that ought to be plenty."

It was plenty—once I stopped trying to find the Kii Threes' limits and just paid attention to the music they made.

Kal Rubinson

[Emlin]

Yeah, but how do they sound?!

[Emlin]

1 minute ago, Ralf11 said:

can the DSP fully correct for shoving them back into the corners?

According to the specs, yes.

[Emlin]

 l want to explain that room acoustics doesn't change based on which speakers are playing.

 

If you put half of the energy into the room whilst maintaining the same SPL at the listening position, you reckon that that's not going to be audible? Plus, via the cardiode, early reflections will be much more benign.

[Emlin]

You'll have to tell us what room you're using. In detail.

[Emlin]

Thanks. I will want full details of your room though. And thanks for your personal opinions, which are just that.

[Emlin]

2 minutes ago, fas42 said:

My approach would be completely different ... I would determine which part of the bottom of the speakers was the most rigid, in at least 4 positions, around the corners obviously in this case, completely ignoring any pads come between the speakers and stand. Then I would work the Blu Tack thoroughly, so that it's as sticky as possible, and apply a major wad of the stuff in those 4 positions earlier determined; the pads have to be completely eliminated from doing anything. And then jam the speaker down hard, really hard on the wads of goo, so it squeezes out.

 

Why do you think that the pads should be eliminated? Is there something wrong with them?

[Emlin]

4 hours ago, fas42 said:

The point of the Blu Tack is to form a certain type of, mechanical, connection between the cabinet and the stand - if the pads are involved they are complicating the link, and may largely counteract the value of using the Blu Tack.

 

This is all open to experimentation - try making Blu Tack the only physical connection, see what it sounds like; then try Blu Tack and pads in combination - have you gained or lost anything by changing to that approach?

 

Ok, I sort of get that, but what's wrong that you think needs correcting?

[Emlin]

2 minutes ago, tomjtx said:

It's not a compass. It's a foldable iPad stand

 

Compass is the model. It's made by Twelve South.

[Emlin]

1 minute ago, Bernstein said:

I have bought the DigiOne Signature, because it has a very low jitter and can be tuned with an uptone Ultracap 1.2, which is way ahead of traditional toroidal LPS. Maybe worth trying? DigiOne beats a lot of other gear times more expensive. 

How low do you think jitter needs to be before you can't hear it?

[Emlin]

4 minutes ago, Bernstein said:

 

What I was trying to say is, that the design very well done with efficiency in mind (using RPI as endpoint feeding I2C) and that the measurements show very good results (there are some reviews which you can find on google). Soundwise it is compared to endpoints in 1000-2000€ range and adding a very good power supply will enhance it further. 

You misunderstood me. My fault, I'm sure, but in any competently designed modern digital component, jitter is already so low as to be inaudible. There is no need to reduce it further.

[Emlin]

2 minutes ago, Emlin said:

You misunderstood me. My fault, I'm sure, but in any competently designed modern digital component, jitter is already so low as to be inaudible. There is no need to reduce it further.

 

Sorry, it was me who was misunderstanding. 

[Emlin]

1 hour ago, AMR/iFi audio said:

 

Yes, these do bring improvements, even not located in the main audio chain. Routers are noisy.

 

In this case, no routers are present, but even if there were, I suspect that all perceived differences would disappear with a proper blind ABX test.

[Emlin]

16 minutes ago, AMR/iFi audio said:

 

*If* one needs ABX in the first place of course. Many people don't.

 

Many people think they don't.

[Emlin]

25 minutes ago, fas42 said:

 

The big problem here is that most ABX tests are so far from "proper" that it isn't funny - the really dumb one is Foobar's, which is relied upon so often - flawed in its implementation so badly that even a desktop's motherboard sound chip makes it obvious that it has issues.

 

Please explain - I'm intrigued!

[Emlin]

1 hour ago, fas42 said:

 

Just a couple of the factors: the key one is that it doesn't use the files it's supposedly playing, no, it resamples to suit itself, and places working copies of that somewhere on the hard drive, possibly fragmented to hell; and, the interface keeps chattering during the replay with progress information ... if one wants to test whether CPU activity is a factor, no chance of eliminating that!

 

I used an extremely basic Nero player at the time, and immediately noticed that the foobar ABX replay was quite degraded compared to that simple software package  - and started to investigate.

 

To get around all that, I use a tin foil hat.

[Emlin]

2 minutes ago, fas42 said:

 

Ummm, it's not good ... if they don't believe that cables improve, or alter, the sound for their gear, but customers are finding otherwise ... what motivation have they to improve the robustness of the engineering, so that in fact it is indeed the case that tweaking is unnecessary ... .

Or tweaking is already unnecessary. There has not been a system built that some will claim couldn't be improved by snake oil.

 

Until they meet blind ABX testing and the supposed differences disappear. Then they claim that they were a bit tired or that their 'golden ears' only work when no-one is testing them.

[Emlin]

1 minute ago, fas42 said:

 

Most blind ABX testing is never organised to a standard that's adequate - there are holes in the procedure that one can drive a truck through ... just doing some test is far short of a test that truly assesses what's going on.

 

Personally, how I address integrity issues is to work backwards - if I had a Kii I would deliberately make life hard for it; feed it very nasty mains power, place sources of high level RF next to them and the cables it's using;  and combine multiple "badnesses", all at once - nothing absurd, but things that could happen in real life. And see what happens ...

 

IOW, find the limits of its engineering robustness - then I have a handle on where it can be improved, .

 

I was of course referring to properly conducted blind ABX testing.

 

You may well wish to play with extreme situations when testing, but that should also be done with properly conducted blind ABX testing.

 

Otherwise your results will be meaningless, and you will have no handle of any sort.

[Emlin]

1 minute ago, fas42 said:

 

My interpretation is that the 6e cable is doing things to get the sound in better shape - more forward is a sign of lesser SQ; cleaner in presentation, focusing a camera are markers of improvement.

 

The takeaway is that the Kii Threes are too sensitive to the quality of that link; firstly because you could hear a definite difference, and secondly because the extra shielding has moved the SQ in the right direction.

 

I look forward to your blind ABX test results with interest. For now, it's just blah.

[Emlin]

On 8/14/2019 at 8:15 AM, firedog said:

I'd prefer this thread doesn't get turned into a cable argument thread. If someone wants to post their personal impressions of cable switching/sound (when using the Kiis), that's fine. But not the back and forth about what cables can and can't do.  It's not a thread about cables. That belongs somewhere else. 
It hasn't happened yet, but if I think the thread goes really off track I'll turn to Chris  as the OP and request removal of posts. 

 

Are you suggesting that people who make extraordinary claims without extraordinary proof should go unchallenged? I know it's not a cables thread, so why let them make it such, with those with scientific doubt silenced?

[Emlin]

Go for it. It may cure you of your condescension. 

[Emlin]

3 minutes ago, George Hincapie said:

 

I had them on home demo. Didn't like them and they went back, so they aren't the be all and end all.

 

 

Would you care to explain why you didn't like them, please?

[Emlin]

3 minutes ago, PleasantSounds said:

 

 

With the prog Q, the width of the equalized band changes somewhat with the gain - this is the "standard" PEQ behaviour.

The constant Q mode has been introduced to ensure that the width of the band remains unchanged irrespective of the gain.

 

Interesting. So was prog Q or constant Q used previously for boundary adjustments? Or are neither involved in that?

[Emlin]

1 hour ago, fpalm69 said:

 I just tried the Chord Hugo TT2 DAC via XLR analog input to the Kii Three speakers.  Yes, it sounded better then the same music file level matched via USB into the Kii Control.  There was a bit more air/space between the musicians giving the impression of better resolution and overall more relaxed and musical.  Not night and day difference but over careful listening over several days switching back and forth I conclude the Hugo DAC sounded better then going through the Kii Control.

So, even with the Kii DSP converting/resampling the Hugo DAC analog signal it still came through as better SQ to my ears.  The standard Kii Control signal path had a fatter bigger soundstage, more forward where instruments tended to blend together.  Through the Hugo is was more reserved, relaxed, everything in it's place with a nice separation between instruments and vocalists.

 

I wonder it the upcoming Kii Preamp will improve the SQ over the Kii Control?  Seems there might be some room for improvement.

I use the Roon Nucleus via USB to the Kii Control, no other streamers or extra devices (I've tried several).  Would like to try the Innous Phoenix USB device.

 

 

Your results, if taken at face value, mean either that the Hugo is not transparent or that the Kii Control is somehow changing the data that is sends to the speakers. I doubt that either is the case.

 

If you were to do your comparisons, this time blind, I strongly suspect that any "differences" would disappear.